Icy super-Earth or Gassy sub-Neptune?

I'm starting to feel like a broken record on this. Headlines are once more trumpeting the discovery of an extra-solar planet—one of well over a hundred that have been revealed within the last decade. This one is apparently not "Earth-like," as a couple others were erroneously described, but rather a "Super-Earth."

Astronomers on Monday announced the discovery of a frigid extrasolar planet several times larger than Earth, orbiting a small red dwarf star roughly 9,000 light-years away.

The finding alters astronomers' perceptions of solar system formation and the distribution of planets in the galaxy, suggesting that large rock-ice worlds might outnumber gas giants like Jupiter.

The newfound planet is about 13 times heavier than Earth and likely has an icy and rocky but barren terrestrial surface, and is one of the coldest planets ever discovered outside of our solar system.

As we have repeatedly seen, though, we can't believe everything we read in the press. So it pays to go to the original material.

Like the planet discovery I wrote about earlier this year, this one was produced by a group of astronomers that have built a system—called OGLE, for Optical Gravitational Lensing Experiment—that monitors millions of distant stars in small regions of the sky, searching for a few cases where intervening objects pass in front. When that happens, the light of the more distant object is bent (thanks to Einstein's general relativity) and focused to cause a temporary brightening—a process known as gravitational lensing. As the invisible foreground object passes by, the background object eventually returns to its original brightness. The pattern of brightening and dimming over time is called the "light curve."

All we know about the planet is based on the shape of the light curve—actually, on the difference between the shape of the light curve and the best-fit model of a single star acting as the lens. In this case, that difference is a few percent at most and lasted a mere couple hours. The observational achievement is impressive and the result of years of effort in constructing the OGLE system.

Astronomers are masters at extracting the maximum amount of information from a minimum of data, and this case is a good example. From the light curve of the entire lensing event, the astronomers make estimates of the mass of the star, the mass of the planet, the separation of the two, and their distance from us. Unfortunately, the process of extracting the information is full of large uncertainties, so the final results are also highly uncertain.

In the scientific paper where the results are written up, the astronomers claim that the 90% confidence interval on the mass of the parent star is 0.2 to 0.72 times the mass of our Sun—an uncertainty of over a factor of three. The distance to the lensing system is also uncertain by a factor of three, meaning that the projected distance of this planet from its parent is uncertain by a similar factor—from a bit over one A.U. (the distance of Earth from the Sun) to over 4 A.U. (not quite Jupiter's distance from the Sun). (Note that this distance serves only as a lower limit to the true separation in space due to projection effects.)

The ratio of the mass of the planet to that of its parent star is uncertain by a factor of two, bringing the total uncertainty in the planet's mass to a factor of seven—from a bit over three times Earth's mass to over twenty times. Uranus and Neptune are the only planets in our Solar System that fit within that mass range (15 and 17 Earth masses, respectively).

Inevitably, but unfortunately, all this uncertainty gets wiped away during the transition from scientific paper to journalistic article. Also inevitably, terms that astronomers use to hype their own results are adopted and used uncritically by reporters.

The lead author of the team that produced this result, Andrew Gould of Ohio State University, repeatedly labels his new find an "icy super-Earth" when quoted by reporters. But in the scientific paper submitted to Astrophysical Journal Letters, that phrase is never used—and "super-Earth" is only used once, in this sentence: "[These two new planets discovered with gravitational lensing] must have a large fraction of rock and ice, but whether these are covered with a thick coat of gas, like Uranus and Neptune, or whether they are “naked” super-Earths such as are theorized to have formed the cores of Jupiter and Saturn, is unclear."

Gould is clearly using the "icy super-Earth" phrase in his public statements to get press attention. An equally (perhaps more) accurate alternative—"Gassy Neptune-like world"—would not attract nearly the same attention. Well over a hundred extrasolar planets have been discovered in the last decade—what's one more without a catchy slogan?

As in this case, good, interesting scientific results can get buried under a mass of speculation and over-enthusiastic extrapolation when communicated to the general public. Scientists do this amongst themselves for fun, to help direct their research in productive directions, and of course to promote their own line of inquiry to their funders. But I suspect this just leads to confusion amongst reporters and readers, and diminishes the public impact and understanding of later, more refined and more accurate, results when they come in.

If NASA's recent decision holds to indefinitely delay the Terrestrial Planet Finder mission—designed to allow us, for the first time in human history, to detect planets roughly the size of Earth at roughly the distance from their parent stars to be conducive to life—perhaps it is because, thanks to all the recent hype, the public already thinks we have the capability to find "Earth-like" planets?